Package 'UPDhmm'

Function to transform a large collapsed VCF into a dataframe, incorporating predicted states along with the log-likelihood ratio and p-value.

Description

Function to transform a large collapsed VCF into a dataframe, incorporating predicted states along with the log-likelihood ratio and p-value.

Usage

addOr(filtered_def_blocks_states, largeCollapsedVcf, hmm, genotypes)

Arguments

filtered_def_blocks_states	data.frame object containing the blocks
largeCollapsedVcf	Input VCF file
hmm	Hidden Markov Model used to infer the events. The format should adhere to the general HMM format from HMM package with a series of elements: The hidden states names in the "States" vector. All possible observations in the "Symbols" vector. Start probabilities of every hidden state in the "startProbs" vector. Transition probabilities matrix of the hidden states in "transProbs". Probabilities associated between every hidden state and all possible observations in the "emissionProbs" matrix.
genotypes	Possible GT formats and its correspondence with the hmm

Value

data.frame containing the transformed information.

---

Apply the hidden Markov model using the Viterbi algorithm.

Description

Apply the hidden Markov model using the Viterbi algorithm.

Usage

applyViterbi(largeCollapsedVcf, hmm, genotypes)

Arguments

largeCollapsedVcf	input vcf file
hmm	Hidden Markov Model used to infer the events
genotypes	Possible GT formats and its correspondence with the hmm

Value

largeCollapsedVcf

---

Function to transform a large collapsed VCF into a dataframe with predicted states, including chromosome, start position, end position and metadata.

Description

Function to transform a large collapsed VCF into a dataframe with predicted states, including chromosome, start position, end position and metadata.

Usage

asDfVcf(largeCollapsedVcf, genotypes)

Arguments

largeCollapsedVcf	Name of the large collapsed VCF file.
genotypes	Possible GT formats and its correspondence with the hmm

Value

dataframe

---

Function to simplify contiguous variants with the same state into blocks.

Description

Function to simplify contiguous variants with the same state into blocks.

Usage

blocksVcf(df)

Arguments

df	data.frame resulting from the as_df_vcf function.

Value

data.frame containing information on the chromosome, start #' position of the block, end position of the block, and predicted state.

---

Calculate UPD events in trio VCFs.

Description

This function predicts the hidden states by applying the Viterbi algorithm using the Hidden Markov Model (HMM) from the UPDhmm package. It takes the genotypes of the trio as input and includes a final step to simplify the results into blocks.

Usage

calculateEvents(largeCollapsedVcf, hmm = NULL)

Arguments

largeCollapsedVcf

The VCF file in the general format (largeCollapsedVcf) with VariantAnnotation package. Previously edited with vcfCheck() function from UPDhmm package

hmm

Default = NULL. If no arguments are added, the package will use the default HMM already implemented, based on Mendelian inheritance. If an optional HMM is desired, it should adhere to the general HMM format from HMM package with the following elements inside a list: The hidden state names in the "States" vector. All possible observations in the "Symbols" vector. Start probabilities of every hidden state in the "startProbs" vector. Transition probabilities matrix between states in "transProbs". Probabilities associated between every hidden state and all possible observations in the "emissionProbs" matrix.

Value

A data.frame object containing all detected events in the provided trio. If no events are found, the function will return an empty data.frame.

Examples

file <- system.file(package = "UPDhmm", "extdata", "test_het_mat.vcf.gz")
vcf <- VariantAnnotation::readVcf(file)
processedVcf <- vcfCheck(vcf,
    proband = "NA19675", mother = "NA19678",
    father = "NA19679"
)

---

HMM data for predicting UPD events in trio genomic data

Description

This dataset provides Hidden Markov Model (HMM) parameters for predicting uniparental disomy (UPD) events in trio genomic data.

states

Five different possible states.

symbols

Code symbols used for genotype combinations.

startProbs

The initial probabilities of each state.

transProbs

Probabilities of transitioning from one state to another.

emissionProbs

Given a certain genotype combination, the odds of each possible state.

Usage

data(hmm)

Format

A list with 5 different elements

Source

Created in-house based on basic Mendelian rules for calculating UPD events.

Examples

data(hmm)

---

Check quality parameters (optional) and change IDs.

Description

This function takes a VCF file and converts it into a largeCollapsedVcf object using the VariantAnnotation package. It also rename the sample for subsequent steps needed in UPDhmm package. Additionally, it features an optional parameter, quality_check, which triggers warnings when variants lack sufficient quality based on RD and GQ parameters in the input VCF.

Usage

vcfCheck(largeCollapsedVcf, father, mother, proband, check_quality = FALSE)

Arguments

largeCollapsedVcf	The file in largeCollapsedVcf format.
father	Name of the father's sample.
mother	Name of the mother's sample.
proband	Name of the proband's sample.
check_quality	Optional argument. TRUE/FALSE. If quality parameters want to be measured. Default = FALSE

Value

largeCollapsedVcf (VariantAnnotation VCF format).

Examples

fl <- system.file("extdata", "test_het_mat.vcf.gz", package = "UPDhmm")
vcf <- VariantAnnotation::readVcf(fl)
processedVcf <-
   vcfCheck(vcf, proband = "Sample1", mother = "Sample3", father = "Sample2")

---